Using a Cluster-Based Method for Controlling the Aggregated Power Consumption of Air Conditioners in a Demand-Side Management Program

Thermo-storage units such as air conditioners (AC) or space heaters offer significant potential for the demand-side regulation and balancing the consumption with the generation. This makes them attractive resources to mitigate the fluctuation and intermittency of the renewable resources, such as solar. This paper presents a control strategy to adjust the thermostat set points of the air conditioners in a way that the aggregated power consumptions of the ACs would follow a desired trajectory, while maintaining customers' comfort through the level of thermostat set points. We used a clustering technique over the time-series power consumptions of individual ACs to find similar patterns. Similar power profiles maybe indicative of similar room temperatures; and hence, adjusting the set point of similar ACs may better maintain customers comfort. Using a mathematical model for the air conditioners, a simulator was built to assess the performance of this controller in different scenarios where the effect of changes in the ambient temperature was also studied. Our results show that this system can follow the desired aggregated power within 10 minutes, which can be used for ancillary services, such as 10-minutes spinning reserve, in power electric systems.